Patients

The study population consisted of all patients with clinical diagnoses of ischemic CVA at our hospital between December 2009 and January 2011. Specifically, we included patients aged >18 years, of both sexes, referred from the neurology clinic with clinical diagnoses of recent (<7 days) ischemic CVA confirmed by brain computed tomography or magnetic resonance. Exclusion criteria were obvious potential cardioembolic sources (atrial fibrillation, mitral stenosis, and intracardiac thrombus). Because the purpose of this study was to test efficacy of the Valsalva maneuver during TEE, we used TTE to preselect patients who were able to perform the Valsalva maneuver; inability to perform the Valsalva maneuver during TTE because of cognitive or coordination impairment was a criterion for exclusion.

Comprehensive TTE (Aplio, with harmonic imaging capability; Toshiba, Tokyo, Japan) was undertaken to exclude potential cardiac causes for stroke. Afterward, patients were instructed to perform the Valsalva maneuver, and its efficacy was tested. The echocardiographer placed a hand on the patient’s abdomen to check for abdominal muscular contraction when acquiring images and while the efficacy of the Valsalva maneuver was assessed by Doppler. We used a 20 cm/sec decrease in transmitral early (E) flow velocity as an objective criteria for adequate strain phase of the Valsalva maneuver. This criterion has been described as an adequate preload reduction during the assessment of diastolic function. However, because this maneuver creates sufficient intrathoracic pressure to decrease preload, it is probably sufficient to also increase the right atrial–to–left atrial pressure gradient during the release phase of the Valsalva maneuver. In addition to this Doppler assessment of preload change and the palpation of muscular abdominal contraction, we also looked for leftward deviation of the atrial septum during imaging. We found that abdominal contraction could be used as a means of judging the effectiveness of the Valsalva maneuver, because it could be assessed during both TTE and TEE, while atrial septal bulging was better observed during TEE. Contrast injections were administered up to three times during the Valsalva maneuver, with images obtained from the apical four-chamber view. A complete transesophageal echocardiographic examination followed TTE on the same day, using the same echocardiographic equipment and a 5-MHz multiplane probe. After oropharyngeal anesthesia, midazolam (0.05 mg/kg) was used for conscious sedation in an intravenous bolus of 1 to 2 mg before and during TEE. Blood pressure, pulse oximetry, and the electrocardiogram were monitored. A search for other cardiac sources of emboli (assessment of the aorta and left atrial appendage) was systematically undertaken. Images of the interatrial septum were obtained from the best imaging plane for septal membrane visualization, typically 50° to 60°. The Valsalva maneuver was then attempted during TEE, simultaneous with abdominal strain assessment. In case the Valsalva maneuver was ineffective, forced abdominal compression and cough were used to complement the examination. At least three saline contrast injections were undertaken during TEE; additional contrast injections were administered with images obtained from other TEE planes if the site of microbubbles passage in the septum was not clear. Contrast injections consisted of 1 mL blood, 1 mL air, and 8 mL saline agitated with a three-way stopcock with two connected syringes, one filled with saline, and administered intravenously from an antecubital vein. Contrast was injected during the strain phase of the Valsalva maneuver, and normal respiration resumed as the first bubbles appeared in the right atrium. A PFO was judged to be present when microbubbles were seen in the left chambers within three cycles after full opacification of the right chambers. Quantification of left atrial opacification was regarded as grade I (mild; one to five microbubbles), grade II (moderate; six to 20 microbubbles), or grade III (important; >20 microbubbles).

Statistical Analysis

Data are expressed as mean ± SD or as percentages. Student’s t test was used to compare groups, and χ ² tests were used for categorical variables. Agreement was tested using κ statistics. Selected digitally captured images were reviewed by two different observers for microbubble presence and quantification. In case of discordance, a third observer’s opinion was requested. P values < .05 were regarded as significant. The study was approved by the ethics committee of the institution; patients gave written consent for the examination.